Automata Learning

Latest papers

15 papers1 paper in the last four weeks, against 1 the four weeks before. 0.0% of all new papers.

Neurosymbolic Routing for Reliable Reasoning on Resource-Constrained Edge Devices
Sep 24, 2026Avyay Sadhu, Alvaro Velasquez, Lekai ChenAutomata Learning Neuro-Symbolic Reasoning
independent researcher · Department of Computer Science, University of Colorado Boulder, Boulder, CO 80309 USA
Temporal-Causal Inference for Reinforcement Learning via Automata Learning
Sep 7, 2026Jan Corazza, Daniil Kaminskyi, Simon Lutz +4Reinforcement Learning Automata Learning
Research Center Trustworthy Data Science and Security, TU Dortmund University, Dortmund, 44227, Germany. · School for Engineering of Matter, Transport, and Energy at Arizona State University, Tempe, Az 85281, USA.
Learning Canonical Register Automata over Ordered Data Domains
Aug 19, 2026Yong Li, Qiyi Tang, Di-De YenAutomata Learning Active Learning
Key Laboratory of System Software (Chinese Academy of Sciences), Beijing, China · Institute of Software, Chinese Academy of Sciences, Beijing, China · School of Computer Science and Informatics, University of Liverpool, UK
ATLAS: Discovering Agent Strategies through LLM-Guided Abstraction and Automata Learning
Aug 14, 2026Ignacio D. Lopez-Miguel, Andreas Happe, Jürgen Cito +3Automata Learning LLM Agents
TU Wien Vienna, Austria · TU Graz Graz, Austria
Learning to Reason with Curriculum II: Compositional Generalization
Jun 26, 2026Nived Rajaraman, Audrey Huang, Miroslav Dudik +3Curriculum Learning Automata Learning
Microsoft Research. · University of Illinois Urbana-Champaign. · Microsoft Research, New York, NY.
Space-Efficient Language Generation in the Limit
Jun 24, 2026Nicolas Flammarion, Chirag Pabbaraju, Hristo Papazov +2Automata Learning Streaming Algorithms
EPFL, Switzerland · Stanford University, USA
Can Agents Infer Environment from Interaction? Evidence from Agentic Automata Learning
Jun 15, 2026Reef Menaged, Gili Lior, Shauli Ravfogel +2LLM Agent Evaluation Active Learning
The Hebrew University of Jerusalem · New York University · Google Research
Synthesizing POMDP Policies: Sampling Meets Model-checking via Learning
May 14, 2026Debraj Chakraborty, Anirban Majumdar, Prince Mathew +2Formal Verification Automata Learning
Nanyang Technological University, Singapore · Tata Institute of Fundamental Research, Mumbai, India · Université Libre de Bruxelles, Brussels, Belgium +1
SMT-Based Active Learning of Weighted Automata
May 8, 2026Tiago Ferreira, Kevin Batz, Alexandra SilvaActive Learning Satisfiability Modulo Theories
University College London, London, UK · Cornell University, Ithaca, NY, USA
Active Reward Machine Inference From Raw State Trajectories
Apr 8, 2026Mohamad Louai Shehab, Antoine Aspeel, Necmiye OzayAutomata Learning
Robotics Department, University of Michigan, Ann Arbor, MI · Independent Researcher · Electrical Engineering and Computer Science Department, University of Michigan, Ann Arbor, MI
PAC learning PDFA from data streams
Apr 2, 2026Robert Baumgartner, Sicco VerwerAutomata Learning Finite-State Machine
Department of Software Technology, Delft University of Technology, Delft, The Netherlands
Emergent Neural Automaton Policies: Learning Symbolic Structure from Visuomotor Trajectories
Mar 26, 2026Yiyuan Pan, Xusheng Luo, Hanjiang Hu +2Automata Learning Visuomotor Policy Learning
Robotics Institute, Carnegie Mellon University, Pittsburgh, PA 15213, USA
Logic-Gated Time-Shared Feedforward Networks for Alternating Finite Automata: Exact Simulation and Learnability
Mar 20, 2026Sahil Rajesh DhayalkarDifferentiable Logic Automata Learning
Arizona State University
Warm Starting State-Space Models with Automata Learning
Mar 5, 2026William Fishell, Sam Nicholas Kouteili, Mark SantolucitoAutomata Learning State Space Models
Columbia University · Yale University · Barnard College, Columbia University
Error-awareness Accelerates Active Automata Learning
Feb 25, 2026Loes Kruger, Sebastian Junges, Jurriaan RotAutomata Learning Active Learning
Institute for Computing and Information Sciences, Radboud University, Nijmegen, the Netherlands

Sep 24, 2026J/K move · Enter open · S save

Neurosymbolic Routing for Reliable Reasoning on Resource-Constrained Edge Devices

Avyay Sadhu, Alvaro Velasquez, Lekai Chen

independent researcher · Department of Computer Science, University of Colorado Boulder, Boulder, CO 80309 USA

Automata Learning Neuro-Symbolic Reasoning LLM Routing On-Device Language Model Inference Numerical Reasoning in Language Models

Open paper AskPDF arXiv

Running a language model on edge hardware provides private and low-latency reasoning without a network connection, and yet the small models that fit on such devices are unreliable on the tasks computers are expected to handle well, such as arithmetic, algebra, and formal logic problems. We argue that much of this unreliability is avoidable. Many queries appearing to demand reasoning are in fact structurally deterministic and permit fast and exact symbolic solutions. Therefore, forcing a probabilistic model to approximate them sacrifices accuracy and energy for little benefit. We present a neurosymbolic router that classifies each incoming query and dispatches it to the cheapest correct solver, sending structured tasks to deterministic engines and reserving the small language model (SLM) for open-ended word problems. Instead of hand-coding the routing logic, we learn a deterministic finite automaton (DFA) with the L* grammatical inference algorithm, using the SLM as a membership oracle and labeled data as an equivalence oracle. On a Raspberry Pi 4B (8 GB RAM, no GPU), evaluated on 100 untested prompts from DeepMind Mathematics, GSM8K, and RuleTaker, learned routing attains 100% routing accuracy and 98.3% overall accuracy with a 512-token reasoning budget (93.3% on word problems), compared with 72.0% for the strongest agent baseline, Program-of-Thought, and 58.7% for a tool-calling agent given the same solvers. Since formatted queries never reach the model, the router answers them in 1-11 ms and, in its 30-token configuration, runs 8.8x faster and 2.8x more energy-efficient than Program-of-Thought.